Cochrane systematic reviews and co-publication: Dissemination of evidence on interventions for ophthalmic conditions

Background: Systematic reviews of interventions provide a summary of the evidence available on intervention effectiveness and harm. Cochrane systematic reviews (CSRs) have been published electronically in the Cochrane Database of Systematic Reviews (CDSR) since 1994, and co-publication (publication of a Cochrane review in another journal) has been allowed since that time, as long as the co-publishing journal has agreed to the arrangement. Although standards for co-publication were established in 2008, the frequency of co-publication and adherence to the standards have remained largely unexamined. Our objective was to examine the frequency of co-publication of Cochrane Eyes and Vision Group (CEVG) reviews, adherence to the co-publication policy, the relative numbers of citations of the two modes of publishing, and differences in times cited in CSRs with and without a co-publication. Methods: We identified all CEVG reviews published by May 30, 2014 in The Cochrane Library. Using keywords from the title, author names, and "Cochrane Eyes and Vision Group", we searched Google Scholar, Web of Science, Scopus, and PubMed databases to identify possible co-publications. We also emailed contact authors of all identified CEVG reviews to ask them whether they had published their CSR elsewhere. We compared each co-publication to the corresponding CEVG review for adherence to the Cochrane Policy Manual (dated June 10, 2014). We recorded the number of times each CEVG review and each co-publication had been cited by others according to Google Scholar, Web of Science, and Scopus, as of June 11, 2014. Results: We identified 117 CEVG reviews;19 had been co-published in 22 articles. Adherence to Cochrane policy on co-publication was moderate, with all authors complying with at least one of four requirements we addressed. Co-publications were cited more often than the corresponding CEVG reviews; CEVG reviews with at least one co-publication were cited approximately twice as often as CEVG reviews without a co-publication. The number of citations varied considerably depending on whether the CEVG review had a co-publication or not. Conclusions: The findings support encouraging co-publication while maintaining the primacy of the Cochrane systematic review. Support for co-publication may be tempered by other factors such as the possibility that CEVG reviews with a co-publication covered more clinically important and timely topics than those without a co-publication. Assuming that citations are a valid measure of dissemination effectiveness, the 15-year CEVG experience with co-publication of systematic reviews suggests that Cochrane authors should be encouraged to co-publish in traditional medical journals.