Keeping up with the literature for librarians means reading in "double-time": not only keeping up with the latest in the health sciences but also keeping up with our ever-changing profession. Below is an annotated list of a few recent articles of interest on clinical decision making, and e-books and e-journals in library and information science. If this list whets your appetite for more, you may want to subscribe to Current Cites, a monthly review of articles of interest in information technology. Subscription information is available at http://sunsite.berkeley.edu/CurrentCites/mailinglist.html.
Clinical decision making
Ingui, Bette Jean; Rogers, Mary AM. Searching for clinical prediction rules in MEDLINE. Journal of the American Medical Informatics Association 2001; 8(4):391-397.
In the tradition of the JAMA Users' Guides to the Literature and the research of Haynes, Wesson and Laupacis et al, Ingui and Rogers conducted a study to obtain an "optimal search filter" for retrieving clinical prediction rules in the literature. Ingui and Rogers conduct a comparative, retrospective analysis of the literature, searching the Annals of Internal Medicine, BMJ, JAMA, New England Journal of Medicine, Annals of Emergency Medicine and the Journal of General Internal Medicine first by hand (the "gold standard") and then using the search filters they derived. The time period they searched in the journals was from 1991-1998. The data set were separated into half: January - June 1991-1998 became the derivation set and July - December 1991-1998 became the validation set. Using the content-analysis software WordStat (a module of SimStat), the authors analyzed word frequency of record titles and abstracts and subject headings. From an initial set of 694 search filters, the authors found that depending on the user's intent (need for high level of sensitivity or a high level of sensitivity and specificity combined) there were specific filters that could assist in the search process. The filter with the highest specificity is: "predict$ OR clinical$ OR outcome$ OR risk$"; the greatest sensitivity was "predict$.ti. AND rule$." (Searches were conducted using Ovid software.) The authors included tables of the useful search filters for clinical prediction rules.
Tierney, William M. Improving clinical decisions and outcomes with information: a review. International Journal of Medical Informatics 2001; 62:1-9.
While much of this material is familiar ground to health sciences librarians, the review article provides a brief summary of the current issues of medical informatics with respect to clinical decision-making. Particularly, Tierney notes that the efforts to create an electronic medical record system capable of supporting clinical decision-making are still an ongoing task. Similarly, he notes that implementing practice guidelines are often hindered by the lack of 'incoming' clinical data. Tierney points out that in order to make guidelines applicable to individual patients, "algorithms with explicit definitions of clinical states and branch points" must be written and, in turn, clinical data must be fed into the system. Ideally, data from the electronic medical record would "record symptoms (and their severity) along with providers' judgements [sic] of a therapeutic intervention's success." While much progress has been made, Tierney concludes that "An effective and affordable integrated system is still a dream for most clinicians." Three tables in the article are of particular interest: a table on "data currently available electronically" [in clinical settings]; one titled, "outcomes to assess in evaluating innovations in medical informatics" and "methods of assessing the effects of innovation in medical informatics."
E-Books: their future, their pricing and libraries
O'Leary, Mick. New academic information model bypasses libraries. Online July/August 2001; 25(4):72.
A brief discussion of the digital book vendors, XanEdu , Questia and ebrary, and the business models under which each one operates. Two of the companies are building digital book collections that can serve an undergraduate population and the third, XanEdu, is repackaging familiar library databases. O'Leary sounds a cautionary note about these companies' business models as each market directly to the end-user -- students and faculty -- and leaves out the library. XanEdu puts a new search engine -- the ReSearch engine -- on ProQuest databases and ABI/Inform. O'Leary finds nothing particularly new about this product though he points out the irony of students paying for services that would very likely be available for free at their library. The more innovative product that XanEdu offers is CoursePacks. CoursePacks is offered directly to faculty, allowing them to choose from a collection of articles and business cases or to create their own online digital CoursePack. The service is free to faculty but once selected for a course is then sold to students at $10 - $25 each. Questia has a collection of approximately 50,000 undergraduate liberal arts titles (none in science, engineering, technology or business). O'Leary found that Questia had a "good recent titles .. on every topic" but beyond that found the collection "disappointingly thin." The collection is fully searchable, however, and priced competitively (including "a $14.95 for 48 hours" procrastinator's special). The third company reviewed, ebrary, (not to be confused with NetLibrary, which markets to libraries), also has a collection of searchable undergraduate titles. The difference is in its payment model, where users pay only for the portion of the book or article that they want instead of buying the whole thing. In summary, O'Leary find the "appeal of these services is not in their content, which is unimpressive by the standards of the smallest academic library, but in their exceptional convenience." And while O'Leary recognizes the very real threat that such services pose to libraries, he predicts that the companies may ultimately find it more profitable to work with libraries to in turn provide access to these digital collections to students and faculty.
Snowhill, Lucia. E-books and their future in academic libraries. D-Lib Magazine 2001; 7(7/8). (http://www.dlib.org/dlib/july01/snowhill/snowhill.html)
Snowhill offers a librarian's perspective of the e-book market. This overview article is a nice counterpoint to the Jasco article below. Snowhill details the findings of the University of California's California Digital Library's Ebook Task Force. The article focuses less on the vendors and more on the larger issues that affect libraries such as software and hardware standards and protocols, digital rights management, access, archiving and privacy. The author points out that at present there is "no established standard for an interoperable e-book format for commercially produced e-books." Vendors are using HTML, XML or PDF formats. The author identifies the standards being developed by Open Ebook Forum (OEB) as the "most promising." These standards would ensure that a given ebook would be operable on both PC and portable reading devices. Another issue is that of digital rights management system (DRMS). DRMS is essentially the enforcement factor in the use of ebooks. DRMS can limit user, time, impose a fee and/or control the extent of content. Currently there are limits on interlibrary loan and classroom use. The major DRMS are: ONIX, Adobe Acrobat Web Buy, XrML (joint effort between Microsoft and Xerox) and Open Digital Rights Language. The author also surveyed 15 academic libraries that were known to be providing ebooks access.
E-Journals in Library and Information Science
Jasco, Peter. Electronic shoes for the Cobbler's children: treatment of digital journals in library and information science for databases. Online July/August 2001; 25(4):46-52.
Jasco takes a look at the coverage of digital library and information science journals in databases. Specifically, he examines the coverage of free digital-only (no print counterpart) LIS journals, choosing ten to assess coverage in databases. The digital journals selected for coverage review were: