CHAPTER 5

Indexing, Abstracting, and Digital Database Resources

LAURA ABATE

As information resources on the Web continue to expand, evolve, and interconnect, health sciences indexes and databases follow suit. MEDLINE, a key dataset of research articles in the health sciences, expanded its coverage by making more of its historical backfiles available online, while PubMed, a popular and freely accessible access point for MEDLINE data, grew in scope to include some additional item types including prepublication records, out-of-scope citations, and articles from the freely accessible archive PubMed Central.1 CRISP, previously useful for identifying research that had been funded but not necessarily completed, evolved into NIH RePORT and NIH RePORTER, which provide access to NIH-funded research projects and the resulting publications and patents along with a suite of tools designed to analyze NIH research activities.2^–3 PDQ, once a self-contained repository of cancer research and treatment information, transformed into a dynamic and more accessible resource by publishing its information summaries on National Cancer Institute webpages and by integrating and enhancing clinical trials information obtained from ClinicalTrials.gov with additional information useful to cancer researchers and patients.4

Health sciences indexes and databases reflect the range of information needed by health-care professionals, researchers, and students, as well as health-care consumers. Consumer health resources (see chapter 11) provide accessible information at a range of reading levels and in a variety of languages to support patient education on health and disease topics. Online textbooks (e.g., AccessMedicine, STAT!Ref) provide health sciences students and health-care professionals with background information, allowing searchers to explore the scope of information available and documenting health sciences knowledge at the time of publication. Where textbooks provide foundational and background information, point-of-care information tools (see chapter 16) are concise, condensed information tools designed to allow health-care professionals to quickly locate, assess, and apply information for patient care. Research databases (e.g., PubMed, Scopus, etc.) provide access to the continuously expanding record of scientific knowledge by documenting and describing individual studies and experiments to support both knowledge acquisition and further scientific inquiry.

While health sciences indexes and databases reflect the range of information needs, they can also be used as a prism for viewing the research process. New information is first available via journal articles that are indexed and abstracted in research databases. The new information may be confirmed or refuted by additional research (and journal articles), and subsequently included in a compendium of information available on a topic and published in a textbook. If the information is strengthened by additional studies and relevant to clinical care, it will be incorporated into a point-of-care information tool. Using health sciences indexes and databases, it is possible to follow information from its first appearance in the scientific literature to its current application. As scientific publications follow the long-standing practice of citing information sources, health sciences databases and indexes can trace the use of information both forward and backward in time via resource citations.

In addition to using health sciences indexes and databases to trace the evolution and aggregation of scientific knowledge over the past sixty years, specialized health sciences databases can go beyond what is published. Using online tools, searchers can identify ongoing research projects as well as those that haven’t yet resulted in publication. Searchers can also identify historical materials including both published documents and items from the grey literature and can increasingly access these materials online as digital repositories increase their holdings and new repositories are established. Health sciences indexes and databases are also being deployed in new ways to contribute to the research process by supporting researchers in the identification of potential collaborators, selection of publications for their research, and creation of novel paths for research via translational research or expansion into new disciplines.

This chapter describes the research databases and foundational resources that are most important to health sciences students, researchers, and practitioners. Resources will be described both in terms of their unique contribution to health sciences research and how they interconnect and relate to additional information tools and knowledge sources. This chapter complements information available and the online resources described in chapter 16, “Point-of-Care and Clinical Decision Support Resources,” and chapter 11, “Consumer Health Sources.”

Contemporary Practice

As indexes have evolved from print to electronic tools and database searching has changed from librarian-mediated to being accomplished by the individual, user expectations have shifted. Many of today’s users cannot fathom an information world where only an “expert” conducts searches. Most users and librarians have grown accustomed to or only known an environment in which searches are conducted by the individual. Users may still rely on librarians for assistance in getting started on a project; for advice on locating materials which are proving difficult to identify; and for systematic and comprehensive searches, but few will request searches primarily through a librarian. In 2007, De Groote, Hitchcock, and McGowan published data showing that mediated searches declined from 2,157 searches in 1990–1991 to 18 searches in 2004–2005.5

As part of their experience as information consumers, today’s users have high expectations for the information tools at their disposal. Both users who have accessed the Web from their earliest years and those who began using it later in their educational or professional careers may have a faulty understanding regarding information availability and the information resources landscape. The seamless delivery of full text via open-linking systems, whether a user is searching PubMed or Google Scholar, may lead users to believe that all information is free and freely accessible. This perception may leave users ill-informed regarding the scope and breadth of information available and of the value of the resources available from health sciences libraries. The rapid retrieval provided by typing keywords into a Web search tool may make it difficult for users to understand the use, impact, and possible advantages of indexing schemes versus full-text searching. Finally, users’ ability to harness the breadth of the Web to find some information on virtually any topic may leave users poorly prepared to identify the best information resource for their needs and to evaluate information retrieved for quality and application.

As the onus of searching has shifted from the librarian to the individual user, the role of the librarian has changed. While librarians remain expert searchers due to their knowledge of information tools and their highly honed skills with individual resources, librarians must now redouble their efforts as instructors to enable users to locate information effectively and efficiently. In their role as instructors, librarians should educate users on the content and prospective uses of information, the structure and design of the data resources, and the search strategies that work best for each resource. This education will occur in a variety of settings according to the library environment, availability of professional staff, and needs of the users. Educational sessions will vary from informal to formal settings (e.g., from Reference Desk inquiries to prepared remarks at Grand Rounds), from general topics to specific resources (e.g., from locating background information to searching Health and Psychosocial Instruments), and one-on-one interactions to classes ranging from a dozen to several hundred students.

While users expect immediate and direct access to electronic information resources, librarians must remain knowledgeable and skilled in locating information in print resources. Some indexes remain accessible only in print format; the information available in print indexes may not be accessible through online search tools. From 1996–2010, the National Library of Medicine completed a project to add older material from Index Medicus and Cumulative Index Medicus to MEDLINE, thereby extending coverage back to 1946.6 While this change added twenty years of data to the MEDLINE database, indexing data from MEDLINE’s various precursors extends even further back, into the nineteenth century. Users who rely exclusively on searching MEDLINE online may miss relevant information that is available in print indexes.

As database coverage changes and electronic resources evolve, librarians need to maintain their knowledge of both print and electronic indexes in order to select the most appropriate research tools. Librarians should monitor coverage dates for relevant databases as the coverage years continue to change, and be able to identify corresponding print indexes that can be used when searches need to be extended to earlier years. Librarians are also advised that comprehensive literature searches will likely require the use of print indexes. In the fourth edition of this book, Perry, Howse, and Schlimgen produced an extensive table describing key print abstracts and indexes, which serves as a valuable reference tool for identifying relevant print indexes.7 In the third edition of this book, Boorkman provided additional detail on the history of medical indexes, including the titles and coverage years of MEDLINE’s predecessors.8 For detailed information on print indexes and their coverage, readers should refer to these editions.

In addition to monitoring the coverage in terms of year spans, librarians should also monitor the content coverage of online indexes. The specific titles and content indexed by databases change over time, and new resources may offer subject-oriented portals to publications. Librarians should be cognizant of the breadth of the indexes and databases which they select, including the journal titles included and whether those journal titles are indexed comprehensively, selectively, or in some combination of those approaches. Comprehensive searches, such as those conducted for systematic reviews, may require careful searching of multiple databases.

Over the past several years, federated search systems and Web scale discovery systems have been launched by health sciences libraries. These search systems frequently provide a single box search interface similar to popular general Web search engines (e.g., Google, Bing) with the goal of providing a straightforward interface for searching across the library’s print and online collections. Federated search systems (e.g., 360 Search, ExLibris’s MetaLib, etc.) function by searching multiple disparate resources (e.g., the library’s OPAC plus licensed and freely accessible databases) in real time. Web scale discovery systems (e.g., EBSCO Discovery System [EDS], SerialsSolutions’s Summon) access a master index composed of data previously harvested from internal sources (e.g., the library’s OPAC and/​or local repositories), existing databases (e.g., PubMed, CINAHL, etc.), and external scholarly publishing platforms (e.g., Science Direct, HighWire, etc.). Both types of systems are highly customizable so the specific content and coverage of a particular system will likely vary from library to library. As with other indexes and databases, librarians should be aware of the breadth and depth of the available search system and be ready to select an alternative resource depending on the needs of the user and the scope of the search.

From Indexing and Abstracting to Full Text

As the electronic environment has evolved, users’ perceptions about databases have shifted. Initially, literature databases mirrored their previously published print counterparts and described articles by author, title, and topic. Early online users were grateful for the convenience and speed with which a literature search could be conducted to identify all articles by a particular author or to identify all articles on one or more topics. However, while early users appreciated the newfound convenience of online searching, users and librarian expectations changed rapidly. Users were no longer satisfied with a printout of citations and abstracts, but began to seek and expect immediate online access to full-text articles.

In contrast to previous decades when print indexes provided the key to locating pertinent full-text articles, contemporary databases offer a range of full-text content. Indexing and abstracting resources such as MEDLINE still echo their print counterparts by recording who/​what/​where data about journal articles. However, these indexing sources are now frequently integrated with linking systems, which allow users to jump from indexing information describing who wrote an article, what it is about, and where it can be found, to a full-text copy of the original article. Link resolvers such as 360 Link and SFX maintain databases of information about a library’s electronic journal titles and their coverage dates, and serve as an intermediary, connecting an article’s record in an online index to the appropriate electronic full-text copy of the article. Link resolvers are also commonly used in full-text databases to connect users from a reference list citation to a full-text copy of the document.

In contrast to the previous example where users continue to search an indexing and abstracting source and link from that database to a full-text document, some databases provide all or part of their content in full-text format. For example, Clinical Key provides access to a range of document types including journal articles, books, practice guidelines, and synoptic clinical care information. Much of the information in ClinicalKey is available in full-text format, while indexing and abstracting information alone persists in other areas. ClinicalKey users will find that the search system pulls results from the ClinicalKey’s extensive full-text holdings as well as from MEDLINE data for which the full-text availability will vary depending on the collections of the particular library. For some articles in ClinicalKey, indexing and abstracting data alone is available and searched, while for other articles ClinicalKey may retrieve a match based on more extensive indexing of the full-text document as well as the indexing and abstracting data.

Continuing on the spectrum of full-text availability are databases that are full text in their entirety. Databases in this category include resources with the online equivalent of traditional textbooks such as AccessMedicine or STAT!Ref, collections of full-text topic reviews such as the Cochrane Database of Systematic Reviews, and full-text point-of-care databases such as DynaMed and UpToDate (see chapter 16). While items from these databases may be indexed in traditional indexing and abstracting sources (as is the Cochrane Database), more frequently, these databases stand on their own as individual knowledge sources.

Core Research Databases

Core research databases form the backbone of medical research. These databases index and abstract original research articles as they are published in scholarly journals. As such, core research databases contain the latest information available by publishing research reports (e.g., clinical trials results), recommendations (e.g., practice guidelines), and opinion pieces (e.g., letters, editorials), and news articles. The core research databases described in this chapter are integral to comprehensive information searches as they provide an ongoing record of scientific information as it is made available to the public.

Core research databases are generally citation and abstract databases, though they nearly always provide some mechanism for linking to full-text documents. Indexing systems and standards vary by database, but some generalizations can be made. Indexing usually contains some subject classification system which can range from an informal collection of author-provided keywords to a highly structured, consistently applied thesaurus of terms. Core research databases also frequently provide additional information about the article, including type of publication (e.g., Meta Analysis, Peer-Reviewed Journal Article), language, and date, as well as information about the research subjects (e.g., human, rabbit), patient characteristics (e.g., gender, age, ethnicity), and research locale (e.g., Taiwan, California).

MEDLINE is the premier biomedical research database in the world. It is produced by the U.S. National Library of Medicine, which provides free access to MEDLINE via the PubMed Web interface. The National Library of Medicine also leases MEDLINE data to commercial database vendors and websites who may provide free access to MEDLINE, or who may sell access to MEDLINE via proprietary search interfaces. Even as the availability of information on the Web and the number of databases has proliferated, MEDLINE has maintained its position as likely the most important research tool in the health sciences and serves as a baseline for health sciences research against which other literature databases are measured.

MEDLINE is highly valued as a research resource for a number of reasons: breadth of medicine and health sciences topics covered; quality standards applied to journals selected for indexing; expanded year coverage; variety of languages and geographic span of journals indexed; a consistent and evolving indexing scheme; and wide accessibility. MEDLINE indexes about 5,600 publications, primarily scholarly journals.9 MEDLINE covers medicine and health sciences broadly including “life sciences, behavioral sciences, chemical sciences, and bioengineering needed by health professionals and others engaged in basic research and clinical care, public health, health policy development, or related educational activities.”10

Journals selected for MEDLINE indexing undergo a stringent review process through which they are evaluated for scope and coverage, quality of content, quality of editorial work, production quality, audience, and types of content, and geographic coverage.11 In 2010, MEDLINE coverage expanded to go back to 1946 by digitizing and incorporating data that was previously available in print: Index Medicus and Cumulative Index Medicus.12 Although MEDLINE is produced by the United States National Library of Medicine, it covers journals published worldwide. Currently, MEDLINE includes journal articles written in thirty-nine languages, and historically has included articles in sixty languages.13

Journal articles are indexed for MEDLINE using a controlled vocabulary called MeSH (Medical Subject Headings). The power of MeSH indexing lies in its consistent application and ongoing adaptation to new concepts and terminology in health sciences. MEDLINE’s consistent and detailed indexing can be used to retrieve and sort items to answer specific research questions. Indexers assign MeSH terms to describe the topics of each article, the publication type (e.g., controlled clinical trial, editorial), and geographics (e.g., continent, region, country, state, etc.).14 MeSH terms can also contain qualifiers or subheadings to further specify the type of information contained in an article (e.g., Economics, Epidemiology, Enzymology, etc.).15 Multiple MeSH terms may be applied to describe different aspects of an article. For example, an article on assessing a chickenpox vaccination program in Brazil might contain the following MeSH terms:

Part of MEDLINE’s influence as a research tool undoubtedly lies in its broad accessibility. Free access to MEDLINE is provided via PubMed (http://www.pubmed.gov/), and numerous additional free and fee-based sources of MEDLINE are available.

PubMed is a popular tool for searching MEDLINE as it is freely accessible and offers a flexible and powerful search environment. The terms PubMed and MEDLINE are frequently used synonymously, although more strictly speaking MEDLINE refers to the dataset, which is available via numerous free and fee-based search interfaces, while PubMed refers to the search interface offered by the National Library of Medicine to search primarily MEDLINE data. PubMed is easily navigable by novice searches, and can also be harnessed by experienced users for complex searches. PubMed can be customized by individual users via “My NCBI” personal accounts, and by libraries or other institutions to reflect local search and display preferences. PubMed customizations include the ability to reflect preferences for full-text linking and local holdings, preferred search filters, and default display formats.16 Via “My NCBI” accounts, users can store searches and citations. PubMed also offers additional search tools including Clinical Queries and other search filters, a single citation matcher, and the ability to make connections with information in other NLM databases (e.g., MedlinePlus, TOXNET).

In addition to MEDLINE data, PubMed contains records for “In Process” citations, which are generally waiting for the addition of MeSH terms. PubMed also contains several other types of articles, which are not in MEDLINE, including citations from MEDLINE journals that precede that journal’s indexing for MEDLINE; additional citations from journals where only selected articles are indexed for MEDLINE; citations to journals that submit full-text articles to PubMed Central; citations to manuscripts of articles written by NIH-funded researches; and citations to book chapters and books available on NCBI Bookshelf.17

CINAHL, the Cumulative Index to Nursing and Allied Health, is a bibliographic database that provides access to the nursing and allied health literature going back to 1981.18 While MEDLINE covers health sciences broadly, CINAHL focuses specifically on the numerous and varied disciplines of nursing (e.g., Anesthesia Nursing, Gerontologic Nursing, Private Duty Nursing, etc.), as well as allied health-care fields, including athletic training, physical therapy, physician assistants, nutrition and dietetics, and respiratory therapy, among others.19^–20 Presently, CINAHL indexes approximately 3,000 journals,21 a significant proportion of which are not picked up in MEDLINE or Scopus.22

CINAHL’s primary focus is journal articles, but unlike MEDLINE, CINAHL also indexes other sources of information. In CINAHL, users can find references to books, audiovisuals, pamphlets, software, dissertations, and research instruments. In CINAHL, the “Publication Types” field describes the format of the original item (i.e., book, journal article, pamphlet, etc.), as well as aspects of information within that document. For example, research results may be indexed to multiple entries under publication type to reflect that the publication was a journal article, and also that it included tables or charts and included a questionnaire or scale. CINAHL also uses several unique publication types that are useful to nursing and allied health personnel including Clinical Innovations, Care Plan, Accreditation, Legal Cases, and Research Instrument.23

CINAHL is indexed using CINAHL subject headings, which are modeled after MeSH.24 CINAHL’s indexing terms include a subject heading to which a subheading can be added in order to specify the type of information presented on a specific topic (e.g., Emergency Nursing/​Legislation and Jurisprudence). While many of CINAHL’s subject headings reflect its focus on nursing and allied health literature and so echo the terminology used in these professions, other portions of the CINAHL thesaurus are drawn directly from MeSH, including disease and drug topics.25

While CINAHL is primarily a bibliographic database, some full-text publications are incorporated. The selected full text in CINAHL includes research instruments, critical paths, standards of practice, government publications, and patient education materials, among others. EBSCO Publishing also provides additional versions of CINAHL with varying levels of full-text and access and indexing of additional titles: CINAHL with Full-Text, CINAHL Plus with Full-Text, and CINAHL Plus.26

Stay updated, free articles. Join our Telegram channel

Join