Retracted systematic reviews continued to be frequently cited: a citation analysis.

BACKGROUND AND OBJECTIVES

To survey the citations of retracted non-Cochrane systematic reviews (SRs) in scientific literature.

METHODS

We searched the Web of Science and Google Scholar from their inception to 30 April 2020 to find the citations of 153 previously identified retracted non-Cochrane SRs. We calculated the numbers of citations before and after retraction separately. We also described how the citation addressed the retraction and how it was used in the article.

RESULTS

A We identified 954 citations of 128 retracted SRs. The number of retracted SRs and citations reached the peak in 2014 and 2016, respectively, and the majority of the citations (n = 580, 60.8%) were in articles published after the SR was retracted. The mean number of citation per retracted SRs was 7.5. 2.6 before and 4.5 after the publication of the retraction notice. Twenty-nine (5.0%) citations indicated the retraction of the SRs in the reference section. Nine of these citations supported the retracted SR's results, and 15 disagreed with them.

CONCLUSION

Retracted SRs continue to be cited after the publication of the retraction notice. Standardized methods are needed to guide the management of retractions and avoid inappropriate citations of retracted articles.