“Rubber stamping” retracted papers

The paper by Budd et al. [1] on article retraction is a valuable addition to the literature on this topic. In preparing an article on publishing ethics in the biosciences, I have been struck by the fact that, in all the medical libraries I have visited, papers that have been formally retracted have not been identified as such in the “hard copy” on the library shelf. This situation appears not to be peculiar to the Antipodes. The majority of medical libraries have been found not to have policies and procedures for alerting readers about retractions [2, 3].

Regrettably, from time to time in medicine, published articles are retracted, for reasons ranging from honest but “lethal” errors in data gathering to underlying fraudulent research. These papers mislead readers, embarrass editors and reviewers, and may cause researchers to head in directions they would not otherwise head. To minimize further damage, it is imperative that the retracted status of papers is readily apparent to the unsuspecting reader or researcher. Thankfully, computer databases tag a retracted paper. For example, MEDLINE uses the words “Retracted Publication” as a subject heading alongside the offending piece. Fortunately, too, the journal itself will issue a brief notice of retraction, “withdrawal of aegis,” or apology in a subsequent issue. However, readers and researchers do not source all articles via computerized searching nor scour journals for notices of retraction. For example, a recent survey at our hospital library suggests that about 50% of articles requested from external libraries are sourced by readers using noncomputerized means. One solution is that all medical libraries, at regular intervals, appropriately stamp the first page (and possibly all pages) of those articles that have been formally retracted. This task would not be onerous; as Budd et al. reported [4], in the period 1966 to 1996, there were 235 papers retracted among all journals indexed by MEDLINE.

Of course, the library of the future may arguably be a bank of computers, devoid of all hard copy, making this chore redundant. For now, I am very interested to learn about the practice of other libraries in their handling of retracted papers. On the basis of my experience Down Under, I believe that libraries may need to be more active in this matter, so that “bad science” is not perpetuated.