Indian Journal of Radiology Indian Journal of Radiology  

   Login   | Users online: 4788

Home Bookmark this page Print this page Email this page Small font sizeDefault font size Increase font size     

 

 Table of Contents    
COMPUTERS IN RADIOLOGY  
Year : 2011  |  Volume : 21  |  Issue : 3  |  Page : 162-169
Using PubMed in radiology: Ten useful tips for radiologists


Department of Continuing Education, QMed Knowledge Foundation, Mumbai, India

Click here for correspondence address and email

Date of Web Publication24-Sep-2011
 

   Abstract 

PubMed contains a bibliography of articles published in around 4800 journals. It combines MEDLINE and OLDMEDLINE (articles from 1960, going back till the 1940s). PubMed is updated on a daily basis; to include both published and ahead of print references. As a radiologist, one can use PubMed to track several journals, track topics, search for specific topics, verify incomplete or incorrect references, store one's own publications, and save selected references; one can also create filters depending on one's own search needs for some regular topics. This article provides some key background knowledge on searching PubMed and also describes some features that are often left unexplored. The PubMed site has undergone many changes in the last few years and this article will update users on the current features.

Keywords: Information retrieval; literature search; PubMed

How to cite this article:
Sriganesh V. Using PubMed in radiology: Ten useful tips for radiologists. Indian J Radiol Imaging 2011;21:162-9

How to cite this URL:
Sriganesh V. Using PubMed in radiology: Ten useful tips for radiologists. Indian J Radiol Imaging [serial online] 2011 [cited 2019 Oct 23];21:162-9. Available from: http://www.ijri.org/text.asp?2011/21/3/162/85362

   Introduction Top


PubMed, at http://www.pubmed.gov, is a bibliographic database compiled by the National Library of Medicine, USA. A bibliographic database is one that incorporates only the bibliographic details of articles of journals (usually a set of journals chosen by set criteria) and not the full articles. PubMed indexes around 4800 journals in medicine and the health sciences. [1]

One can search PubMed for a list of articles by topic, by author or by journal. One can also search for combinations of these and limit searches in several ways. PubMed itself does not provide the full articles, but will link each article to the publisher's website and to the article itself, if it is available. Whether the article is available for free or for a price is decided by the publisher.

PubMed is usually searched for information based on research or studies published in journal articles. In the context of patient care, one may search for systematic reviews, meta-analyses, guidelines, and randomized controlled trials, so that one gets evidence-based information. With internet access being available almost anywhere, PubMed can be now accessed from clinics, wards, radiology centers, or homes; it can even be accessed through mobiles, using 'PubMed Mobile.'


   Using PubMed - Common Errors Top


PubMed is a very large database and so the search process works more efficiently if a structured approach is used. Here are the two most common errors that inexperienced/novice searchers commit:

  1. Using only key words/phrases to search: A key word search yields far too many references and includes lots of irrelevant hits. Searching PubMed with key words (or phrases) like 'HRCT in bronchiectasis,' 'fetal ultrasound' or 'MRI spine in lymphoma' will result in a search for references that contain these words/phrases even though the articles may or may not be about these topics.
  2. Looking only for full articles: PubMed can point to the most relevant articles on the search topic. If one looks only for articles available free online one may miss very important articles with valuable information. Do remember that many "paid" articles may be available in local libraries.



   Personalize PubMed Before Starting to Search Top


Personalizing PubMed is an investment that will take only a couple of minutes. Follow these simple steps to create a 'MyNCBI' account [Table 1]:
Table 1: Steps in personalizing PubMed

Click here to view



   Ten Useful Tips for Searching PubMed Top


1. Searching using 'fields'

Every bibliographic record has fields. These are like column headings in tables. One can search for a word or phrase only in one or more specific fields. Commonly searched fields are 'Author,' 'Title,' 'Title/Abstract,' and 'Journal Name.' [2],[3]

A few examples and contexts for field searches are illustrated below:

a. If a quick reading list of articles on 'perfusion MRI for brain tumors' is needed, search by typing:

brain[ti] tumors[ti] perfusion[ti] MRI[ti]

The references obtained will have all four words in the title [Figure 1]. Furthermore, one can choose one or more relevant articles and click 'Related citations' to find more articles.
Figure 1: A Title Search gets you articles with your search terms in the title. Add [ti] after each term. This is good for a quick reading list of articles on a topic

Click here to view


b. For a list of articles by an author, search by using the field [au].

Wilson [au]

If a search is made without [au] after 'Wilson,' additional articles on Wilson disease, Wilson's test, etc. will be retrieved.

c. For a list of articles in a journal, use [jour]

Radiologic Clinics of North America [jour]

The other fields that can be used for efficient searching are: [TIAB] for title and/or abstract, [DP] for date (year) of publication, and [PL] for place (country) of publication.[Figure 2]
Figure 2: A MeSH Search uses terms that are standardized. A search is done by choosing the correct MeSH term for every search. In the example shown, the ensuing search will pinpoint and display several references about 'ultrasound diagnosis of brain tumors'

Click here to view


2. Combining concepts

If one wishes to search for more than one concept, the words 'AND,' 'OR,' and 'NOT' can be used. These words must be in upper case.

AND means the resulting references must have all of your search terms.

OR means the resulting references may have one or more of your search terms.

NOT is used to eliminate a term from the results.

Some examples and contexts for combining concepts are given below:

  1. Magnetic resonance imaging[tiab] AND ultrasound[tiab]: This will retrieve articles that contain both terms in the title and/or in the abstract.
  2. Magnetic resonance imaging[tiab] OR ultrasound[tiab]: This will retrieve articles that contain the terms 'magnetic resonance imaging' or 'ultrasound,' or both, in the title and/or in the abstract.
  3. Magnetic resonance imaging[tiab] NOT ultrasound[tiab]: This will get articles that contain only 'magnetic resonance imaging' and will eliminate articles containing 'ultrasound.' When NOT is used, articles that contain both ' magnetic resonance imaging' and 'ultrasound' will be eliminated. Thus, a search is also performed with AND in order to select important references from that list. The terms AND, OR, and NOT are known as Boolean operators.


3. Searching for references using [tiab] alone is inadequate

Searching for a topic that is available only in the title and/or abstract is inadequate because of two possible scenarios. Examples highlighting this inadequacy are given below:

  1. The topic 'words' may be present in the title and/or abstract, but the article may not be about them. To illustrate, a search for 'acute traumatic lesions' may fetch an inappropriate article titled 'Bone marrow edema pattern around the knee on magnetic resonance imaging, excluding acute traumatic lesions.'
  2. The title of an article may not be explanatory and the reference may not have an abstract. As an example, let us examine the following reference: Tudorica A, Thomas CR Jr, Huang W. Invited commentary. Radiographics. 2010; 30:716-9;


The full text of this journal is available at http://radiographics.rsna.org/content/30/3/716.full and tells us that the article is about MRI for tumors. But if a [tiab] search for either of these terms is performed, PubMed will not fetch this reference as neither of the words appears in the title and as there is no abstract.

4. Searching topics using MeSH terms

In order to search for articles about terms or phrases, the search should be performed using MeSH terms. MeSH stands for 'medical subject headings.' These are very specific and standardized terms used to describe every article in PubMed. Searching using MeSH terms will fetch highly relevant references. MeSH terms are added to every bibliographic record by indexers who read the full article and use MeSH terms to 'describe the article.' Searching with MeSH terms will fetch articles about the specific term, whether the term appears in the title/abstract or not. An example of a search without using MeSH term is shown here to highlight the inherent inconvenience of that approach. An article titled 'Sonographic evaluation and pregnancy complicated by diabetes' does not have the words 'gestational diabetes' in the title or abstract although it is about ultrasonography in gestational diabetes. Nevertheless, if a search is performed for gestational diabetes [MeSH], the above quoted article will appear within the retrieved list in the search results.

Searches using MeSH term are more focused and meaningful. [Table 2] illustrates this point while searching for two random topics: 'ultrasound diagnosis of brain tumors' and 'MRI spectroscopy for hepatic steatosis.' MeSH terms are stored in the MeSH database. MeSH terms are standardized, and choosing the best and correct MeSH terms for every search is the single most important step. On the PubMed home page, the link to the MeSH database is on the lower right-hand corner.
Table 2: Searching PubMed using MeSH [Figure 2]

Click here to view


Two additional options are available for advanced MeSH searches. These are located below the subheadings for a MeSH term, namely 'Restrict to MeSH major topic' and 'Do not include MeSH terms found below this term in the MeSH hierarchy.' Checking either of these will narrow down a search result. The first option gets articles where the search term has very major coverage and is not just a small part of the article. The second excludes the lower 'specific terms;' for example, when the searching for MR spectroscopy, terms like electron spin resonance spectroscopy and nuclear magnetic resonance will be excluded.

5. Limiting the searches

After running any search you can see a link just above the search box - 'Limits.' Clicking this, one finds several options to limit the searches.

6. Finding review articles

There are two kinds of review articles: regular and systematic. Regular review articles are overviews on a topic and are written by experts. They are obtained in any search by clicking the filter 'Review' on the right side of the results page. Systematic reviews are usually written by a group of experts after identifying large studies and analyzing them. These maybe found in the Cochrane Database of Systematic Reviews ( www.thecochranelibrary.com ) or in journal articles through PubMed.

To search for systematic reviews of a topic (e.g., diagnosis of stroke by MRI), use the following strategy:

Stroke/diagnosis[major] AND magnetic resonance imaging[mesh] AND (Cochrane Database Syst Rev[jour] OR systematic review[ti])

This retrieves either Cochrane reviews or systematic reviews in journals on the topic.

7. Saving your searches and tracking them

After performing any search, one can save the strategy and get updates on that topic. Logging into MyNCBI is mandatory for this. The 'Save search' button is just above the search box. One can set up email alerts to automatically receive new articles whenever they are added to PubMed. Significantly, one can set up an alert for more than one journal by creating the strategy described below and running a search

AJNR[jour] OR Radiologic Clinics of North America[jour] OR Indian journal of Radiology and Imaging[jour]

After running the search, clicking the 'Save search' button enables one to choose various options to receive email alerts as and when new issues are added to PubMed [Figure 3].
Figure 3: Journal Alert. Logging into My NCBI is mandatory for saving a search and setting up email alerts to automatically receive new articles from one or many journals whenever they are added to PubMed

Click here to view


Searches can also be saved for topics like the ones described earlier. Setting up alerts is an option; it is not compulsory for every saved search. One can change settings for 'Saved searches' by clicking the My NCBI link.

8. Saving specific article references

After performing a search or receiving an email alert, you may wish to save one or more of these references for use in the future. Saving is possible by selecting and clicking the check box/es of the reference/s. Clicking the 'Send to' link (below the search box) shows various options. Choosing the option 'Collections,' allows one to create a new collection or add references to an existing collection on the topic [Figure 4]. The collections can be shared by choosing to make them 'Public' and storing the URL.
Figure 4: Send to Collections. Saving specific article references is possible by selecting and clicking the check box/es of the reference/s. Clicking the 'Send to' link displays the options. Choosing the option 'Collections' allows one to create a new collection or add references to an existing collection on the topic

Click here to view


Other value-added features include the options of a) saving bibliographies of articles one has written by choosing 'Send to My bibliography'; b) adding non-PubMed citations by going to My NCBI and then clicking 'My bibliography' followed by 'Add citation', and finally, choosing the appropriate type of citation and adding the details.

9. Free articles

When a search is performed, a link 'Free full text' is seen at the right side of screen. Clicking this link gives a list of free articles on a topic.

In addition, there are a few hidden sources of free articles. One of the main such sources is 'author manuscript' in PubMed Central, an archive of journal articles. In PubMed, type in the following PMID - 19521011 and search. PMID is the unique number for every article/citation in PubMed. In the result page, there are two icons: 'IOP' and 'Free author manuscript' in the bibliographic record. By clicking IOP, which is the publisher's icon, payment for the article will be required. The other icon will give the author's final manuscript which very often serves the purpose.

Besides, there are journals that are freely accessible in India due to assistance given to many developing countries. These will not be marked free, but clicking the publisher's icon works nevertheless. An example is the journal Annals of Internal Medicine.

A few Indian journals have not been provided icons in PubMed. The Journal of Association of Physicians of India, National Medical Journal of India, Indian Journal of Medical Ethics, and many others have full articles on their respective websites, but have not requested PubMed to provide the free icon links. If one finds an important article in any Indian journal, it would be wise to go to the journal website and check if the article is available online

10. Filters in PubMed

A typical search results page shows two filters on the right side: All, Review, and Free full text. One can create up to 15 filters by clicking 'Manage filters' located just below the three above-mentioned filters. To do this one needs to be signed into MyNCBI.

On the right side of the 'Manage filters' page, one can see four 'categories.' The first is 'Popular,' for selecting any filter one may want to use. The second, 'Linkout,' allows one to select various linked resources as filters. One useful option is to search for 'India' using the search box. A list of libraries show up; selecting the check boxes under the 'Filter', and 'Link' icon allows one to check the availability of articles from the listed libraries.

The third category, 'Properties,' allows creation of filters for articles, for example, describing certain age-group populations, language of publication, etc. The fourth category - 'Links,' permits linking to other databases and certain select citations. Of special interest are the two given below (these can be retrieved by searching for 'PMC' in the search box)

  1. The 'Links to PMC' option will create a filter to select articles available free through PubMed Central [Figure 5].
  2. The 'Links to PMC references,' option creates a filter that will choose a subset of articles. For each article in the subset filter, if one clicks the title it shows which PubMed Central article has cited the particular article [Figure 6].
    Figure 5: Filters. There are several filter options through 'Links' which permit linking to other databases. The 'Links to PMC' option will create a filter to select articles available free through PubMed Central

    Click here to view
    Figure 6: Filters. The 'Links to PMC references' filter chooses a subset of articles. For each article in the subset filter, if one clicks the title it shows which PubMed Central article has cited that particular article

    Click here to view


    A custom filter for radiology journals can also be easily created. In the 'Manage filters' page, there is a link 'Create custom filter' (in the upper half, towards the center). On clicking this link, one is asked to enter a name (e.g., 'My radiology journals'). In the 'Search' box, the strategy needs to be entered (e,g., AJNR[jour] OR Radiologic Clinics of North America[jour] OR Indian journal of Radiology and Imaging[jour] ). Then if one clicks 'Run filter,' the number of results is displayed. The 'Save filter' button saves the filter and lists it under 'Your PubMed filter list' [Figure 7]. Ensuring that the check box next to the filter is selected is important if it is to show up as a filter when one uses PubMed.
Figure 7: Radiology journals. Creating a custom filter for radiology journals: From the 'Manage filters page,' clicking 'Create custom filter,' a folder 'My Radiology Journals' and then by running and saving a search strategy, the custom filter is listed under 'Your PubMed filter list.' Selecting the filter allows the filter to be visible in PubMed

Click here to view



   Conclusion Top


PubMed is updated every day with new records and is thus a goldmine for medical literature. This article has covered some of the most important tips and features for using PubMed. With constant use one can keep discovering more features and tips.

The MeSH database and indexing of each article with these terms is the unique feature of PubMed and makes a world of a difference in searching for relevant articles. The libraries filter is another unique feature that needs to be incorporated by all medical and radiology libraries. As of the moment it is lamentable that only about eight libraries in India have uploaded their journal holdings (list of journals, with years of coverage) to enable the library filters to be efficiently used. It is important that every library uploads its holdings so that filters of libraries can be created.

 
   References Top

1.PubMed. Available from: http://www.pubmed.gov/. [Last accessed on 2011 Jun 14].   Back to cited text no. 1
    
2.Sriganesh V. Searching PubMed and other databases. Indian J Dermatol Venereol Leprol 2005;71:139-42.  Back to cited text no. 2
[PUBMED]  Medknow Journal  
3.Sriganesh V. PubMed -a beginner′s guide to searching biomedical literature effectively. J Postgrad Med Educ Train Res 2008;3:9-13.  Back to cited text no. 3
    

Top
Correspondence Address:
Vasumathi Sriganesh
A-3, Shubham Centre, Cardinal Gracious Road, Chakala, Andheri (E), Mumbai - 400 099
India
Login to access the Email id

Source of Support: None, Conflict of Interest: None


DOI: 10.4103/0971-3026.85362

Rights and Permissions


    Figures

  [Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7]
 
 
    Tables

  [Table 1], [Table 2]

This article has been cited by
1 Google and PubMed for physicians: How to find information without getting lost [Google y PubMed para médicos: ¿cómo buscar información sin perderse?]
Muñoz Núñez, C.F. and Sendra Portero, F.
Radiologia. 2013; 55(SUPPL.1): S37-S46
[Pubmed]
2 A study on pubmed search tag usage pattern: Association rule mining of a full-day pubmed query log
Mosa, A.S.M. and Yoo, I.
BMC Medical Informatics and Decision Making. 2013; 13(1)
[Pubmed]
3 Google y PubMed para mdicos: cmo buscar informacin sin perderse?
C.F. Muoz Nez,F. Sendra Portero
Radiologa. 2013; 55: 37
[Pubmed] | [DOI]
4 A Study on Pubmed Search Tag Usage Pattern: Association Rule Mining of a Full-day Pubmed Query Log
Abu Saleh Mohammad Mosa,Illhoi Yoo
BMC Medical Informatics and Decision Making. 2013; 13(1): 8
[Pubmed] | [DOI]



 

Top
 
 
  Search
 
   Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
    Email Alert *
    Add to My List *
* Registration required (free)  


    Abstract
   Introduction
    Using PubMed - C...
    Personalize PubM...
    Ten Useful Tips ...
   Conclusion
    References
    Article Figures
    Article Tables

 Article Access Statistics
    Viewed4623    
    Printed216    
    Emailed1    
    PDF Downloaded654    
    Comments [Add]    
    Cited by others 4    

Recommend this journal