Qualitative and Quantitative Measures of Indexed Health Sciences Electronic Journals

E-JOURNALS AND ONLINE INFORMATION

Ann C. Weller, MA Context Little is known about qualitative and quantitative characteristics of indexed health sciences electronic journals (e-journals). LECTRONIC PUBLISHING HAS THE Methods To determine peer-review practices and qualitative and quantitative charac- potential to offer many advan- teristics of different types of indexed health sciences e-journals, 3 types of e-journals in- tages over traditional print pub- dexed in MEDLINE were compared (type 1, completely electronic with no print counter- lishing, but concerns over qual- part; type 2, print and electronic versions with the same title but each publishing some ityE exist.1 Several studies have identified unique content; and type 3, print and electronic versions containing equal content). the following qualitative and quantita- Results There were 13 type 1 journals, 16 type 2 journals, and 16 type 3 journals. tive measures that are appropriate for Most journals in each category (85%-94%) imply or state the use of peer review. Sig- Ͻ electronic journals (e-journals): peer- nificant differences (P .05, analysis of variance) exist among the e-journals for the review practices,2 citation patterns,3 inclusion of complex types of publications (clinical trials, randomized controlled trials, meta-analyses, and practice guidelines) (15%-100%), editorials (0%-75%), letters to structured abstracts,4 and editorial 5 the editor (10%-88%), and case reports (17%-94%); the average number of items boards. The objectives of this study indexed in MEDLINE (22.5-544.5); and the number of complex publication types, edi- were to determine if there are differ- torials, letters, and case reports. ent editorial peer-review practices or Conclusions Type 1 e-journals do not have the qualitative or quantitative complex- variations in qualitative and quantita- ity of traditional print journals. Although editors’ statements on editorial peer review tive measures among different types of are similar, there are differences in number and type of materials included in the 3 indexed health sciences e-journals. different types of e-journals. JAMA. 2002;287:2865-2866 www.jama.com METHODS There are 3 types of e-journals: type 1 were randomly selected from Abridged tions to authors. Items indexed from are completely electronic with no Index Medicus titles. electronic type 2 e-journals always car- regular print version; type 2 are titled Data for this study were collected from ried a unique identifier in MEDLINE, the same both in the print and elec- MEDLINE and each journal’s Web site usually an “e” preceded the page num- tronic versions, but each publish homepage. All items published in the ber. TABLE 1 lists those e-journal char- some unique content; and in type 3, year 2000 and indexed by August 2001 acteristics that were the same for both both the print and electronic versions were included. No e-journal in this study type 2 electronic and print versions. publish equal content. Each type 2 was selectively indexed by MEDLINE. More than 80% of all e-journals stated version has its own international Qualitative and quantitative mea- or implied the use of peer review and standard serial number, but type 3 sures included: editorial peer-review required original material. These data e-journals have 1 international stan- statements, inclusion of editorial boards, were located by searching the Web site dard serial number. requirement of original research or struc- homepages and reading the instruc- Thirty e-journals were listed as ac- tured abstracts, number of articles cited, tions to authors sections and other in- tive MEDLINE titles in June 2001: 13 the inclusion and number of complex formation about the e-journals. The av- were type 1, 16 were type 2, and 1 was publication types, case reports, editori- erage number of type 1 e-journals that not yet published. Abridged Index als, and letters to the editor. The index required structured abstracts (36%) was Medicus is a subset of 120 MEDLINE terms clinical trials, randomized control less than for either type 2 or type 3. All titles, which is considered a solid, core trials, meta-analyses, and practice guide- type 2 and type 3 and 92% of type 1 set of health sciences journals, and was lines were used for complex types of pub- e-journals listed an editorial board. used in selection of type 3 e-journals. lications. Author Affiliation: Library of the Health Sciences, Uni- To compare type 1 and type 2 e-jour- versity Library, University of Illinois, Chicago. nals with traditional print journals, 16 RESULTS Corresponding Author and Reprints: Ann C. Weller, MA, Library of the Health Sciences, University of Illi- that had electronic access to current is- Type 2 e-journals always shared a Web nois Library, 1750 W Polk St, Chicago, IL 60612 sues and qualified as type 3 e-journals site homepage and had 1 set of instruc- (e-mail: [email protected]).

Downloaded From: https://jamanetwork.com/ on 09/23/2021 INDEXED HEALTH SCIENCES ELECTRONIC JOURNALS

review statements, requirement of origi- Table 1. Journal Characteristics* nal research, and the inclusion of edito- Type, % rial boards. A lack of a statement on peer 1 2 3 review or original research does not mean Journal Characteristics (n = 13) (n = 16) (n = 16) that no policy exists. Editors address the Stated peer-review status statement Definite 77 81 62 editorial peer-review process in journal Implied 8 13 25 guidelines more frequently today than Not mentioned 15 6 13 they did in 1987.2 The structured ab- Require structured abstracts 36 53 69 stract was proposed for medical jour- Mention original research 82 88 87 nals in 1987.4 One third to two thirds of Has editorial board 92 100 100 all types of e-journals in this study used *There were an average 4.62 type 1, 383.06 type 2, and 369.31 type 3 electronic journal articles cited in the year 2000 (P = .01). this format. This study confirmed the ear- lier study by Harter3 that type 1 e-journals continue to be infrequently cited. Table 2. Comparison by Journal Type* This phenomenon may continue until Type 2 (n = 16) Type 1 Type 3 type 1 e-journals become established, MEDLINE (n = 13) Electronic Print (n = 16) mainstream medical literature. Presence of publication types, % Many electronic type 2 publications Complex 15 25 88 100 Editorials 30 0 75 75 serve a specialized function and may rep- Letters 10 13 75 88 resent a trend to publish journals with Case reports 17 38 88 94 different content in the print and elec- Average No. of publications indexed tronic versions. The electronic type 2 Total No. indexed 22.5 48.8 459.3 544.5 publications often include only 1 type Complex 0.5 6.5 50.2 35.1 of feature, such as short reports, rapid Editorials 0.5 0 18.1 11.0 communication, or letters. For ex- Letters 1.5 6.9 31.6 48.4 ample, one electronic type 2 e-journal Case reports 0.2 22.5 47.7 47.9 published all letters electronically, while *PϽ.001 for all comparisons. another published only “ultra-rapid” A search in August 2001 of Institute The type 1 and the electronic type 2 were communications, and another let the au- for Scientific Information’s Web of Sci- less likely (PϽ.001) than print type 2 or thor choose which version for case re- ence determined frequency of citations type 3 e-journals to publish editorials, ports. Eleven (69%) of the 16 type 2 to each e-journal. The ISI captures all letters to the editor, and case reports. journals stated specific criteria for in- cited references in each journal it cov- For the year 2000, type 1 and elec- clusion in the electronic only version. ers independent of whether the title is tronic type 2 published significantly The 13 indexed health sciences type covered by the database; 54% of type 1 fewer items than print type 2 and type 1 e-journals lack the depth and breadth e-journals are covered by ISI. An aver- 3(PϽ.001). There were also statisti- of traditional print journals. The type age of 4.62 articles published in type 1 cally significant differences for the av- 1 and electronic type 2 e-journals pub- e-journals in the year 2000 were cited, erage number of items indexed as com- lished significantly fewer indexed com- whereas more than 350 articles pub- plex types of publication (P=.01). Both plex types of publications, editorials, let- lished in type 2 and type 3 e-journals in type 1 and electronic type 2 published ters, and case reports than either type the year 2000 were cited (P=.01, single- significantly fewer editorials (P=.03), 2 print or type 3 e-journals. Data from factor analysis of variance). letters (P =.002), and case reports this study provide an early snapshot of TABLE 2 compares results when sepa- (P=.004) than print type 2 and type 3. e-journals. The type 1 e-journals will rate data exist for type 2 e-journals. probably become more numerous and There were statistically significant dif- COMMENT more accepted, but they currently do ferences among the e-journals (PϽ.001) The following qualitative measures were not have the complexity of traditional indexed as complex publication types. similar for all types of e-journals: peer- print journals.

REFERENCES 1. Wright SM, Tseng WT-C, Kolodner K. Physician tronic journals: an impact study. J Am Soc Inform Sci. Association. CMAJ. 1994;150:1611-1615. opinion about electronic publications. Am J Med. 2001; 1998;49:507-516. 5. Steffens DL, Robbins JB. The role of editors and 110:373-377. 4. Taddio A, Pain T, Fassos FF, Boon H, Ilersich AL, editorial boards in journal publishing. In: McClure 2. Weller AC. Editorial policy and the assessment of Einarson TR. Quality of nonstructured and struc- CR, Hernon P, eds. Library and Information Science quality among medical journals. Bull Med Libr As- tured abstracts of original research in the British Research: Perspectives and Strategies For Improve- soc. 1987;75:310-316. Medical Journal,theCanadian Medical Association ment. Norwood, NJ: Ablex Publishing Corp; 3. Harter SP. Scholarly communication and elec- Journal and the Journal of the American Medical 1991.

Downloaded From: https://jamanetwork.com/ on 09/23/2021