Strength of the Brandon/Hill list for core medical journals
April 7, 2004
Susan H. Keesee
Graduate Student
School of Information and Library Science
Health Sciences Library
Campus Box # 7585
University of North Carolina at Chapel Hill
Chapel Hill, N.C.27599
Telephone: 919-962-0801
E-mail:
“Each profession regards its own journals important, but few fields have professional journals that are arguably more important to the rest of the community than medical journals.” 1
Librarians, writers of articles, editors, publishers, advertisers and journal readers judge the quality of scholarly journals both objectively and subjectively. For assistance with purchase, continuation and cancellation decisions, librarians weigh a combination of quality indicators as impact factors, circulation figures, manuscript reject rates, and inclusion on consensus core lists.2
The Brandon/Hill core listsare well known to collection development librarians serving professional audiences in the fields of medicine, nursing and allied health. The most recent version (2003) of recommended print books and journals for the small medical library features 141 journal titles, 60 of which are suggested for initial purchase if a library collection was being built from scratch.The 60 journal titles belong to what is generallycalled the minimal core list. Although there is a brief description of the selection process used to compile the list in the introduction to earlier versions, the third paragraph preceding the 2003 list highlights that the authors’ felt the need to defend their subjectivity. 3
“The authors have considered input – pro and con – from health science librarians, editors, clinicians, subject specialists, and vendors. However, the final choice of publications is ours – based on our opinion. From the beginning, the Brandon/Hill selected lists have always been based on the opinion of the compilers. There is no model or scientific method employed. Journal title selections remain relatively stable.”
The compilers also qualify the list by noting that “[T]he majority of the publications are clinical in content … The journal list is not intended to support research activities.” Ibid., 2.
List inclusion as a quality indicator
This study is designed to investigate whether the Brandon/Hill list has maintained its strength as one of the indicators of journal quality at two academic health sciences/medical libraries. Anecdotal evidence on the importance of the list is given,followed by a brief history on consensus lists in the field of medical literature.
TheJournal of family practice(JFP) is listed among the 141 titles butnot on the minimal core list. One of JFP’s editors wrote that because of the subjective methods used to select journals it is impossible to understand the exclusion of that title from the minimal core, citing that it is the oldest and most widely cited journal of original research in the specialty of family medicine. 4
Others in the medical field charge that librarians’ use of traditional quantitative measures and/or inclusion on consensus lists is biased, unscientific, and statistically imperfect. A group of researchers noted that bibliometric parameters such as journal citation rates, impact factors, circulation, and manuscript acceptance rates are controversial and have not been validated as predictors of journal quality. 2 The group developed a 22-item evaluation instrument to help identify which journals published the high quality research articles to allow readers to quickly select the most up-to-date medical literature to review.As summarized by Technicalities columnist Tony Stankus, “low (sic) and behold, our criteria and theirs seem to agree to a great degree!”1
Corelists: a brief history
Access to the world’s biomedical journal literature is provided through Index Medicus and it online counterpart MEDLINE. No library other than possibly the National Library of Medicine can maintain subscriptions to the 4500 titles indexed and included in the MEDLINE database and the 3630 titles in Index Medicus.5
Alfred Brandon built upon the usefulness of the core holdings lists sporadically published by the American Medical Association between 1940 and 1959. His experience building collections at health sciences libraries and his participation in an active medical librarians’ association in Baltimore while director of the Welch Medical Library at the Johns Hopkins School of Medicine made him acutely aware of the need for such a reference and selection tool. The first selection list for small medical libraries was published in 1965.6 The 1960s also saw the heightened interest in the development of an Abridged Index Medicus(AIM) and other core collection lists.7
Dorothy Hill joined the effort of making biennial updates to the Brandonselection list in the 1970s and later the original listhad two spin-offs; nursing in 1979 and allied health professions in 1984, also updated biennially. 6
Another biennally updated work is the Core collection of medical books and journals published by the Medical Information Working Party (MWIP), which strives to offer a more international perspective than the Brandon/Hill lists.8 The MWIP consists of librarians, booksellers, publishers, and periodical agents. Two American reviewers of the second edition,published in 1994,predicted “that as standards and practices span national boundaries more and more in the years ahead, resources such as the Core Collection will have increased relevance to U.S. medical librarians.” Ibid., 524.
A library will use such tools as the Brandon/Hill lists and the Library for Internists when evaluating collections to meet and exceed Joint Commission on Accreditation of Healthcare Organizations (JCAHO)standard IM9.1 which is “The hospital’s knowledge-based information resources are available, authoritative and up to date.”9
Just one tool
Two examples highlight how the list is an essential tool but not the sole arbiter for collection development.
To aid in a needs-based collection development project, a retrospective study on1,224 document delivery requestsfrom rural health professionals included comparisons of the journal titles and subject divisions on the Brandon/Hill lists and in theAIM.10Although core titles from either list did not fill more than 30% of the total requests, approximately 70% of the titles populating the core lists were requested. The list is not gospel, but a guide from which the researchers developed a list of core titles to satisfy their user population’s information needs.
A system of assigning up to 10 points to journal titles was part of Florida State University’s collection development process to define a core list for the first academic medical library to be established since electronic resources mushroomed in the 1990s.11A journal title’s inclusion on the 2001 version of the Brandon/Hill list of books and journals for a small medical library provided two points and a third point for designation as a first purchase title, called the minimal core in this article. Individual points were assigned for inclusion on other lists and for holdings at 10 and 16 of 18 community-based medical school libraries.
Buying power fades
Medical libraries are by no means immune from the intense scrutiny of collections when budget cuts are looming. Because of the stability of journal titles on the Brandon/Hill list, it has been a guide to compare cost increases across time. Some 38 years after publication of the first list, the average subscription has increased from $13.90 to $391.92 – an increase of 2,720%.3Capturing prices for electronic versions is a complex process that compilers of the list do not plan to pursue at this time. While this may be a limitation of the list to some, it shows its strength by focusing on the content of materials, not the format.
Problem Statement
With all their subjectivity and reliance on print, is a consensus core-holdings list such as the Brandon/Hill selected list of print books and journals for the small medical libraryrelevant in an environment of intense cost cuttingand the shift from print to electronic platforms?
Althoughstrength of the core list could have been declared simply by comparing titles to recent publicly available journal cancellation lists from Duke’s Medical Center Library12 and UNC’s Health Sciences Library,13 this would have been a short article and not much information gained about the characteristics of two academic medical libraryjournal collections. None of the titles cancelled by either library were on the Brandon/Hill list.
Methodology
Between March 28 and April 5, 2004, the author searched for current print and electronic holdings for all journal titles on the 2003 Brandon/Hill selected list of print books and journals for the small medical library in the online library catalogs at the University of North Carolina at Chapel Hill (officially abbreviated as UNC-CH, but referred to here simply as UNC) and Duke University.14, 15The ISSN and publication frequency were verified in both catalogs to ensure a match.
If a catalog record existed, the title’s location at either UNC’s Health Sciences Library orDukeUniversity’s Medical Center Library was confirmed. Data confirming current electronic subscriptions at UNC were verified in the E-journal finder.16Duke’s catalog did not have a separate e-journal finding resource. Cancellation dates for print at UNC were verified at the link for Current Journals at HSL.17Because access to Duke’s current journals database is limited to on-campus and Duke-affiliated users, presence or absence of the phrase “currently received” in the catalog holdings record was relied upon for accuracy in this study.
Data were collected in an Excel workbook separating the minimal core titles on one worksheet from the remaining titles on another worksheet. The following expansions and contractions were made to each list prior to analysis:
Although not included on the alphabetic journal title list, The American Journal of Medicine (ISSN 0002-9173) was among those asterisked on the journal list by subject. This adjusted the list to 60 in the minimal core and 81 remaining titles.
The Journal of bone and joint surgery was counted as two titles since there is an American and a British version, each with its own frequency. The Journals of gerontology also had A and B versions with different frequencies and availability, so this title also counted as two. The minimal core list contained 62 items.
Two titles -- the Cumulative Index to Nursing and Allied Health Literature (CINAHL) and Index Medicus -- were excluded from the list of 81 remaining titles.
The author justified the exclusions because both are held in the index areas at each library and are used as reference tools.
Results and Discussion
The data shown in Table 1 provide a very cursory view of the nature of the two library collections when compared to the Brandon/Hill list. Much more information can be gained by analyzing the overlap, print cancellation history, and gaps in coverage.
Minimal core list
When combining both library’s collections, there is complete coverage by either print or electronic format of all 62 titles on the minimal core list. Each library lacks one title in both print and electronic format; very fortunately not the same title. Version B of the Journals of gerontology (ISSN 1079-5014) is no longer available as a current title at UNC. Print was cancelled in November 2003. Duke does not have any holdings (print or electronic) of the Journal of substance abuse treatment (ISSN 0740-5472).
As of December 2003, UNC no longer receives print for 25 more titles on the minimal core list. Electronic access is available for all 25 of these titles. Including Version B of the Journals of gerontology, UNC lacks electronic access to nine titles. Of those nine, Gastroenterology (ISSN 0016-5085) is the only common title Duke does not have access to electronically.
Trusting the accuracy of only the online catalog record, Duke lacks only three titles in print format from the minimal core list. Electronic access is not available to four titles and three of these are titles that can be accessed electronically at UNC.
Remaining core
Patronswith access to both UNC’s Health Sciences Library and Duke’s University Medical Center Library will find no gaps in coverage but should navigate the online catalogs carefully if they have a preference for print or electronic copies. If a patron was willing to accept both print and electronic access, Duke offers all titles in the remaining portion of the core list. UNC does not hold current subscriptions to three titles: Journal of Critical Care (0883-9441); Physical Medicine and Rehabilitation Clinics of North America (1047-9651); and Progress in Cardiovascular Diseases (0033-0620).
Of the 79 studied of the remaining core list, UNC does not currently subscribe to 32 print titles, 29 which are available electronically. Electronic access to 18 titles is not available at UNC, seven for which there is also not electronic access at Duke.
Of the remaining core, Duke does not hold current subscriptions to 20 titles in print and 10 titles electronically. UNC holds print to 12 of the 20 and electronic access to 3 of the 10 titles Duke lacks.
Conclusion on core list strength
Access to current issues of 140 of the 141 titles in the Brandon/Hill selected list for small medical libraries is available at both UNC and Duke.Due tooverlapping coverage in print and electronic formats there are no titles that cannot be accessed between the two libraries. No analysis was attempted to determine if inclusion on the minimal core list affects other quality measures, however, the editors of the Journal of Family Practice should be pleased that through intense cancellations by both universities their title continues to be currently available at both libraries in both print and electronic formats.
Based on this analysis of which titles are currently received at two academic medical libraries, the core list of 141 titlesdemonstrates enduring strength.
Implications for libraries studied
Should the list continue to endure its strength, it is hoped that these two special libraries will continue to make provisions to maintain between them at least one print subscription of all core titles. In atragicsituation when network access is not available to hospital clinicians who must access an item in this core clinical literature, it would be most unfortunate if a paper copy were not available.
Notes
1. Tony Stankus, “The Patient is Going to Make It: Doctors Diagnose Today’s Medical Journals,“ Technicalities 22, no. 4 (July/August 2002): 4, 10-11.
2. Kirby P. Lee, et al., “Association of Journal Quality Indicators with Methodological Quality of Clinical Research Articles,” JAMA 287, no. 21 (June 5, 2002): 2805-8.
3. Dorothy R. Hill, et al., Brandon/Hill selected lists. Levy Library, Mount Sinai School of Medicine, New York University.
Accessed March 24, 2004, and on later dates.
4. Mark Ebell, “The Brandon/Hill list,” Journal of the Medical Library Association 90, no. 1 (January 2002): 111.
5. National Library of Medicine. NLM Fact Sheet on journal selection for Index Medicus/MEDLINE.
6. Dorothy R. Hill and Henry N. Stickell, A History of The Brandon/Hill Selected Lists. Levy Library, Mount Sinai School of Medicine, New York University.
7. Dorothy Dralle, within “Communications to the Editor,” Bulletin of the Medical Library Association 57, no. 4 (October 1969): 402-3.
8. Jonathan Eldredge and Susan Gerding Bader, book review: “Core Collection of Medical Books and Journals,” Bulletin of the Medical Library Association 83, no. 4 (October 1995) 523-4.
9. Prudence W. Dalrymple and Carol S. Scherrer, “Tools of improvement: a systematic analysis and guide to accreditation by the JCAHO,” Bulletin of the Medical Library Association 86, no. 1 (January 1998) 10-6.
10. Josephine L. Dorsch and Victoria Pifalo, “Information needs of rural health professionals: a retrospective use study,” Bulletin of the Medical Library Association 85, no. 4 (October 1997) 341-7.
11. Barbara S. Shearer and Suzanne P. Nagy, “Developing an academic medical library core journal collection in the (almost) post-print era: the Florida State University College of Medicine Medical Library experience,” Journal of the Medical Library Association 91, no. 3 (July 2003) 292-302.
12. Pat Thibodeau and Judy Woodburn, “Journal Review Project,” Duke University Medical Center Library News Issue # 290 (October 2003) 3-6 and linked from UNC-HSL’s journal cancellation website:
13. Health Sciences Library, Journal Collection Faces Major Cancellations. and List of proposed cancellations: Web pages accessed March 24, 2004.
14. UNC-Chapel Hill Libraries catalog,
15. Duke University Libraries catalog, then click on the word “catalog” under “Research Tools”
16. UNC-Chapel Hill Libraries E-journal finder :
17. Current Journals at HSL link:
Table 1. Tabulation of the currently received holdings of titles on the Brandon/Hill selected list of journals for a small medical library at two academic libraries
UNC print / UNC e-access / Duke print / Duke e-accessMinimal core (62 titles) / 36 / 53 / 59 / 58
Remainder of list (79 titles) / 47 / 61 / 59 / 69
Percent of all 141 titles / 59% / 81% / 84% / 90%
More data in Excel spreadsheet, linked as a separate file. Edited with the advice of Selden Durgom Lamoureux, instructor of INLS 226, Managing serials in an electronic age (e-mail = ).