EMIT Foundation Module: Resources

A bibliographic database is an organised collection of references to documents. In this setting a reference is information about the document such as author, abstract, journal title and year of publication.

Medical bibliographic databases are essential search tools for medicines information. The medical databases most likely to be available to hospital pharmacists are MEDLINE and Embase.

Most databases index the journal literature (e.g. clinical trials reports, case reports, systematic reviews) but some also include books, conference reports, newsletters and other forms of publication. Indexing provides a description or identification of a document using a set of rules, vocabulary (thesaurus) and keywords.

As tens of thousands of biomedical journals are published worldwide, an individual bibliographic database will not index all of the literature. Selection criteria are usually based on a number of factors including whether the journal is peer-reviewed, the number of times it is cited in the literature, how long the journal has been established and the language of publication.

You may need to search more than one database to find the information you are looking for. If both MEDLINE and Embase are available to you, you should search both.

SHPA members can access MEDLINE and CINAHL through the EBSCOhost interface. Log-in to the SHPA website, then access EBSCOhost from the Publications and Resources tab on the SHPA website. There is an instructional webinar called EBSCO Resources Training on the SHPA eCPD site.

General tips on searching bibliographical databases

When starting a new search, it’s best to keep your search wide. If you retrieve too many hits, you can always narrow your search.
Many databases operate using a concept called ‘Boolean logic’. If you want to combine two terms in a search, you can use the Boolean operators AND, OR and NOT. For example:

• cats AND dogs will produce articles that mention both cats and dogs in the same article. This is the most commonly used Boolean term.
• cats OR dogs will produce some articles that mention cats and some articles that mention dogs and some that mention both cats and dogs.
• cats NOT dogs will produce articles that only mention cats. Articles that mention cats and dogs will not be included.

Each database will be organised differently. Take time to learn how to use the thesaurus of each database that is available you. This will help make your search more effective.
You may need to use US drug names and spellings if you are using an American system. For example, paediatric (pediatric), salbutamol (albuterol). Some databases will automatically convert US and UK spellings and names.

This section cannot replace hands-on experience. Experiment with as many different databases as you can.

MEDLINE^®

Established in 1966, MEDLINE (Medical Literature, Analysis, and Retrieval System Online) is the US National Library of Medicine’s (NLM) bibliographic database. It contains over 23 million references to journal articles in clinical sciences and biomedical research and covers over 5200 journals.

There are several different ways to access Medline. Many hospitals have access via OVID MEDLINE.

MEDLINE does not always provide the full text of a journal article; in most cases an abstract is provided. If the journal can be accessed electronically there may be a hyperlink to the relevant website; some of these may be free or may be available via your organisation’s subscription. There is usually a lag-time between articles being published and being indexed in MEDLINE.

There is no right or wrong way to search MEDLINE. It can be searched using the thesaurus, an author’s name, a journal title, or a word in the text or title of the article abstract. If you are not looking for a single specific article, use the thesaurus for the best results.

Using the Thesaurus

Using the thesaurus is essentially searching for a subject. The MEDLINE thesaurus is called ‘MeSH’ which stands for MEDLINE Subject Headings. MeSH terms are arranged into ‘trees’, where very broad search terms are the ‘trunk’ of the tree and more narrow terms are the branches and twigs.

In the example shown below ‘Cardiovascular Diseases’ is the broadest term and ‘Shock, Cardiogenic’ is the narrowest term.

 

Exploding Terms

If you explode a MeSH term you will include all the terms listed under your selected term in the thesaurus tree.

Using the cardiovascular disease MeSH tree above:

• if you want to find very broad articles about cardiovascular diseases, that may or may not mention myocardial infarction, then do not explode the term cardiovascular diseases.
• if you want to locate articles about cardiovascular diseases (including those that definitely mention myocardial infarction), then explode the term cardiovascular diseases. The explode facility will include all the specific terms ‘beneath’ it in the tree hierarchy.

When the term that you select is further down the branch of a tree it will narrow your searching.

If you use the PubMed version of MEDLINE it will automatically explode your chosen term.

OVID MEDLINE does not automatically explode terms for you. To explode a term using OVID MEDLINE you will need to tick the ‘Explode’ box.

If you are in any doubt, always explode; you’ve nothing to lose because you can always limit your search later on if too many results are obtained.

Using subheadings

If your search is producing too many hits, consider using a subheading, also called a ‘qualifier’. Subheadings are terms that you can add onto your MeSH term to make your search more specific (e.g. drug therapy, adverse effects, metabolism, pathology).

For example you might want to add ‘drug therapy’ into your search about myocardial infarction to only search for articles about the drug treatment of myocardial infarction.

For most versions of MEDLINE there is a screen where you can choose to add subheadings.

Using limits

In addition to MeSH terms and subheadings, each article is tagged with a variety of terms such as the age group of the population studied, the nature of the participants (e.g. human vs. animal) and the type of article (e.g. review).

If, despite adding a subheading to your search, you still have too many results, consider adding one of these ‘tags’ or ‘limits’.

If your search about the drug therapy of myocardial infarction has produced too many hits then consider adding a ‘human’ limit or the limit for review articles.

Focusing terms

Most MEDLINE records have between 6 and 15 MeSH terms assigned to them. Two or 3 of these are marked as the article’s major MeSH terms, which describe the main topics of the paper. In the MEDLINE record these terms will be marked with an asterisk. In some versions of MEDLINE you can choose to find papers that consider your search term as one of the major topics. This process is called ‘focusing’.

Supplementary concepts

There is a lag-time between the development of new drugs and their entry into the thesaurus. During this time some versions of MEDLINE (e.g. PubMed) refer to them as ‘supplementary concepts’. If the thesaurus indicates that your term is a supplementary concept, it will normally give you a ‘registry number’ to use instead.

Alternatively you can just use your new term anyway, turn the thesaurus off and MEDLINE will search the title or text of an article abstract for it. This is also known as a text word or keyword text search.

Combining Terms

The use of Boolean logic applies to MEDLINE searches and you can combine MeSH headings and supplementary concepts using AND, OR and NOT.

Read the following examples and then try the exercise.

Example 1

Your oncology pharmacist asks you for help in finding recent evidence to support the use of lenalidomide in patients with multiple myeloma.  

• Start by thinking about the MeSH headings you want to check in the thesaurus.
• Go to the NLM thesaurus at http://www.ncbi.nlm.nih.gov/mesh and enter lenalidomide. The thesaurus indicates that lenalidomide is a MeSH heading and gives a small description of the term (‘scope’). By selecting the lenalidomide hyperlink it also gives a list of all the subheadings (qualifiers) that you can choose from. If you scroll further down the page you can see where lenalidomide fits into its trees.
• Now go back to the search page and enter the term for multiple myeloma. The thesaurus indicates that multiple myeloma is a MeSH heading, gives the scope and subheadings and shows where multiple myeloma fits into its tree structure. 

• The subheadings for multiple myeloma include ‘drug therapy’. You may want to consider adding this to narrow down your search. 

So far you have two MeSH headings – lenalidomide and multiple myeloma. You also have one subheading for multiple myeloma – drug therapy.

• Now consider whether you are able to use any limits to narrow your search further. You may want to select only human studies, for example.

You have now constructed your Medline search. These will be the terms you will enter into the version of Medline that you have available to you.

Example 2

A pharmacist asks you to perform ‘a quick MEDLINE search’ about the clinical trial evidence for benralizumab for uncontrolled, severe eosinophilic asthma for a committee meeting in half an hour.

As in example 1, first take time to think about your terms.

• Go to the thesaurus at http://www.ncbi.nlm.nih.gov/mesh and enter benralizumab. At the time of writing the thesaurus indicated that benralizumab was still a supplementary concept and not yet a MeSH heading. It gives you an entry term that you can use instead of a MeSH term. Or you can simply enter benralizumab and run a keyword or text word search.

• Now think about the disease. If you search for eosinophilic asthma you will see that there is a MeSH heading for asthma, but there is no suitable subheading. Eosinophilia is also a MeSH heading.

You can try searching for asthma/drug therapy, eosinophilia/drug therapy. Depending on how may results you get you might combine the terms. Or for a narrower search you might combine eosinophilia/drug therapy AND benralizumab.

For newly marketed drugs you may not want to add limits to your search. In this example if you get too many hits with your first search, consider adding limits for clinical trials.

Medline Exercise

Now it’s your turn. Read the following enquiry and try to find the terms in the thesaurus that will enable you to perform a Medline search.

Your neurology pharmacist calls into your office to find out if peripheral neuropathy has been associated with simvastatin.

What MeSH headings will you use?

Show answer

What subheadings, if any, will you use?

Show answer

Do you need to add limits?

Show answer

What extra information would be helpful to this search?

Show answer

What terms did you use for your search?

Show answer

Embase^®

Although less well known than MEDLINE, in many ways Embase is better for medicines information enquiries. It includes references to over 41 million articles in 8100 indexed journals, including all of MEDLINE and 2900 journals not covered by MEDLINE.

The basic structure is similar to MEDLINE, although search terms may differ between the two databases. The main advantages of Embase over MEDLINE are:

• it covers a larger number of biomedical journals. In particular, it is more comprehensive for pharmaceutical, European and drug-related journals.
• the subheading and limit functions are more comprehensive; for example, you can limit searching to a specific route of drug administration.
• the thesaurus includes more drug names.

PubMed and TRIP

PubMed and TRIP are freely available databases. MEDLINE is a subset of PubMed.

PubMed only uses keyword searches (no subheadings) which may require a different method of searching to obtain a narrow search result.

TRIP contains a subset of PubMed, systematic reviews, controlled trials, patient information, NICE, AHRQ, Cochrane, e-textbooks, evidence-based synopses, guidelines and includes a selection of tools to assist with searching.

A series of online training videos, tutorials and tours are available at PubMed Online Training and TRIP to help you.

Emcare®

Emcare is a nursing and allied health database for practice, research, or education. It includes all nursing specialties, social work, mental health, emergency and trauma medicine and critical care. Emcare includes complete integration with the Emtree thesaurus. Emcare covers more than a thousand journals not covered by Embase and includes some unique records that are not searchable in other nursing or allied health databases.

CINAHL Complete

CINAHL is the Cumulative Index of Nursing and Allied Health Literature. It provides indexing for over 3000 nursing and allied health journals.  The CINAHL subject headings follow the MeSH structure, as used in Medline. The database provides abstracts or full text papers for selected journals. It is helpful for enquiries with a nursing theme, such as nurse prescribing, nurse-led admission clinics etc. 

PsycINFO^®

PsycINFO is a bibliographic database that provides systematic coverage of the professional and academic literature in psychology and related disciplines, including medicine, psychiatry, nursing, sociology, education, pharmacology, physiology, and linguistics. PsycINFO contains more than 5 million records including abstracts from books, journals and dissertations.

Cochrane Library

The Cochrane Collaboration is an international organisation that aims to promote evidence- based medicine. The Cochrane Library is most well known for its Database of Systematic Reviews, which contains protocols and reviews prepared by the Cochrane Collaboration. The Cochrane Library also contains the Cochrane Central Register of Controlled Trials (CENTRAL). In Australia, access to the Cochrane Library is funded by the Federal Government.

Google Scholar

Google Scholar is less precise than MEDLINE and Embase and search results can be inconsistent, although it is sometimes useful for accessing full-text articles and conference abstracts.

In-house archiving databases

Many Medicines Information Services have an in-house database that they use to archive the details of past enquiries. Enquiries are entered into the database and are usually given an individual identifying code and can be keyworded. Examples of software databases used in Australia include MiDatabank and ANDIN.