The continued citation of retracted publications in dentistry

Publications are retracted for many reasons, but the continued use and citation of retracted publications presents a problem for future research. This study investigated retractions in the dental literature to understand the characteristics of retracted publications, the reasons for their retractions, and the nature and context of their citations after retraction.

In September 2018, the authors identified retracted dentistry publications using the Retraction Watch database. Citations to those publications were retrieved from Scopus and Web of Science. Characteristics of retracted publications and their citations were collected, including study design, reasons for retraction, and nature of citation (positive, negative, or neutral). We used chi-square tests to determine if there were notable differences between retracted publications that were cited following retraction and those that were not, and if there were relationships between the nature of the citation, the study design of the original publication, and its reason for retraction.

Of the 136 retracted publications, 84 were cited after retraction. When restricted to English language, 81 retracted publications received citations from 685 publications. Only 5.4% of the citations noted the retracted status of the original publication, while 25.3% of citations were neutral and 69.3% were positive. Animal studies were more likely to be uncited after retraction, while in vitro studies and randomized controlled trials were more likely to be cited. Retracted publications that were cited negatively were more likely to have been retracted due to scientific distortion than those that were cited positively or neutrally. Retracted publications that were cited negatively were also more likely to be observational studies than those cited positively or neutrally.

Retracted publications in dentistry are continually cited positively following their retraction, regardless of their study designs or reasons for retraction. This indicates that the continued citation of retracted publications in this field cannot be isolated to certain research methods or misconduct but is, instead, a more widespread issue.