Advancing PubMed? A comparison of third-party PubMed/Medline tools
| DOI | https://doi.org/10.1108/LHT-06-2016-0066 |
| Pages | 669-684 |
| Published date | 21 November 2016 |
| Date | 21 November 2016 |
| Author | Lorna Elizabeth Wildgaard,Haakon Lund |
| Subject Matter | Library & information science,Librarianship/library management,Library technology,Information behaviour & retrieval,Information user studies,Metadata,Information & knowledge management,Information & communications technology,Internet |
Advancing PubMed?
A comparison of third-party
PubMed/Medline tools
Lorna Elizabeth Wildgaard and Haakon Lund
Royal School of Library and Information Science, University of Copenhagen,
Copenhagen, Denmark
Abstract
Purpose –Systematic review s of biomedical liter ature are used to inform p atient treatment .
Yet the acquisition o f relevant literature is provin g increasingly challenging du e to the large volume
of information that needs to be searched, filtered and collocated. There is a need to improve
the efficiency of biomedical literature searches. PubMed remains the primary resource for
biomedical literat ure, and as PubMed mak es the Medline data and Entrez PubMed Programming
utilities freely available, any developer can produce alternative tools to search the database.
The authors question if PubMed still provides the superior search interface for systematic searches
or if the innovative ness of third-party tools provid e alternatives worth considerin g. The paper aims
to discuss these issues.
Design/methodology/approach –In all, 76 third-party tools that build on PubMed content were
identified in a PubMed search and in published studies known to the authors. Only tools that
provided free access to the broad PubMed content and designed specifically to enhance the search
were included, reducing the set to 16 tools. The functionality of each tool within the scenario of a
systematic search w as compared across 11 a spects. A systemati c search in PubMed was use d as
study control.
Findings –The 16 tools limited rather than advanced the sorting, filtering, and export functionality
required in a systematic search. The reproducibility of the searches in these sources was reduced.
The study shows that PubMed remains the superior provider for searching, identifying, and exporting
biomedical literature for systematic reviews.
Originality/value –The work contributes to the discussion of how librarians can help researchers
navigate the biomedical literature in systematic reviews.
Keywords Bibliographic systems, PubMed, Online retrieval, Information searches,
Search and retrieval, Systematic search
Paper type General review
Introduction
The first step of a systematic search is planning the search and in this step, systematic
review authors can especially benefit from collaboration with librarians (Foster, 2015).
Librarians’assistance surpasses general instruction in how the interface to the
bibliographic database supports the methodological demands to the structure and
scope of the search, the sensitivity of the search, bias and selection of articles.
Our expertise in information and knowledge management facilitates critical appraisal
activities. Rather than applying this expertise for our own professional practice, the
medical librarian has the unique opportunity to inform the library users on the
theoretical aspects of carrying out systematic searches and produce up-to-date critical
appraisals of biomedical search software that are suggested as better ways of
synthesizing medical evidence (Booth and Brice, 2004).
Since biomedical researchers need to interpret an increasing quantity of data
(Islamaj et al., 2009) and the publication rates of scientific articles are exploding, there
Library Hi Tech
Vol. 34 No. 4, 2016
pp. 669-684
©Emerald Group Publis hing Limited
0737-8831
DOI 10.1108/LHT-06-2016-0066
Received 17 June 2016
Revised 8 August 2016
Accepted 22 August 2016
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/0737-8831.htm
669
Comparison of
third-party
PubMed/
Medline tools
are increasing demands on text mining technology to improve the relevance and
precision of searches in the biomedical literature (Coccia et al., 2009). This demand has
led to the development of third-party applications aiming to improve the efficiency of
searching the millions of indexed papers (referred to as citations in the PubMed
terminology) in PubMed (Keepanasseril, 2014; Lu, 2011; Connor, 2010). Medline data
and NCBI Entrez Programming Utilities (the Global Query Cross-Database Search
System that powers PubMed provided by the National Center for Biotechnology
Information) are freely available for developers from academia and industry and allow
the creative development of third-party tools complimentary to PubMed. These tools
are designed to simplify the search modalities of PubMed, thus decreasing the problem
of information overload and syntactic correctness in PubMed searches (Coccia et al.,
2009). Ultimately, they aim to improve the specificity and sensitivity of exhaustive
searches and ensure highly relevant replies to medical queries (Islamaj et al., 2009).
Keepanasseril (2014) and Lu (2011) present reviews of third-party web tools to
access, search and process Medline citations. These reviews presented 28 and 24 tools
respectively, overlap 13, but do not test the functionality of these choosing instead to
list the key features, technical developments and processing capabilities of each tool.
In this paper, we compare the search functionality of third-party web tools to PubMed,
specifically in identifying, retrieving, and exporting biomedical literature as part of a
systematic search process.
The systematic search is a scientific methodology used to create medical sys tematic
reviews. The method sets requirements to the process of the search. The search has to
be exhaustive to capture all the available evidence so that conclusions based on the
literature retrieved in the search are unbiased, it must be methodological, transparent,
the rationale of the search defendable, and importantly, the search must be
reproducible so that it can be repeated and conclusions verified (Lefebvre et al., 2011).
PubMed is traditionally the main database used for searching biomedical literature and
constructing systematic reviews, but to search systematically requires the user learn
how the interface works on par with an information professional (field descriptors,
Boolean logic, combining search blocks, index terminologies, database structure,
mapping and matching functionalities, filtering, limits, etc.). Hence developers of
third-party tools are keen to assist the novice user by simplifying the technicalities of
the search and enhancing user-interactivity through advancing visualization and
manipulation of results.
PubMed: contents
Available since 1996, PubMed includes over 25 million citations (time coverage 1946 to
present) from the biomedical literature. It includes: the Medline database, in-process
records of articles yet to be quality controlled and indexed with medical subject
headings (MeSH), ahead-of-print citations that precede the article’s final publication in a
Medline indexed journal, pre-1966 citations that have not yet been updated to current
MeSH or Medline status citations to full-text articles in PubMed Central (PMC), book s
on the NCBI bookshelf and manuscripts published by NIH-funded researchers
(National Institutes of Health).
PubMed: search interface
PubMed supports block searches, Boolean logic, truncation, masking, and tagged field
searches that limit the search to specific bibliographical fields indexed in the database.
670
LHT
34,4
To continue reading
Request your trial