Redundant publication in the journal Radiology

To retrospectively quantify the incidence of redundant publication in the journal Radiology and to compare the present study findings with those published for other journals and medical specialties.

Two readers estimated the incidence of the redundant publication of original articles in Radiology in the year 2001. Original research articles published in 2001 were analyzed by searching MEDLINE on the PubMed server to identify articles that may have represented a duplication of the original Radiology article. MEDLINE was searched between January 1999 and December 2003 by using the surname and initial(s) of the first author. Potentially redundant articles were identified after similarities in titles and abstracts were analyzed. The full versions of all potentially redundant articles and of the corresponding index articles were then retrieved from the library. The potentially redundant article was then compared with the index article. Criteria for redundant publication were as follows: Compared with the index article, the potentially duplicate article had (a) a similar hypothesis, (b) a similar number of subjects, (c) similar results, (d) at least one author in common, and (e) no or little new information.

In 2001, 362 original research articles were published in Radiology. Two instances of redundant publication were found among these articles, and both were considered to be partially redundant publications due to series expansions (ie, increased numbers of study subjects) of 50% and 52%.

Redundant publication appears to be less frequent in Radiology than in the other journals and specialties for which redundant publication information has been reported.