Background To evaluate the number of citations for Cochrane Methodology Reviews after they have been updated or co-published in another journal.
Methods We identified all Cochrane Methodology Reviews published in the Cochrane Database of Systematic Reviews (CDSR) before 2018 and searched for any versions co-published in another journal in the Web of Science Core Collection database up to 16 August 2022. The included reviews were in two cohorts: those that had been published and updated in CDSRand those that had been published in CDSR and co-published in another journal. The primary outcomes were the number of citations to the updated and the original Cochrane Reviews in the five years after the updated review had been published and the number of citations to co-published and non-co-published Cochrane Reviews in the first five years after publication of the co-published version. The secondary outcome was the ratio of an adjusted impact factor (IF) (excluding the data for the co-published Cochrane Review) and the actual IF of the co-publishing journal.
Results Eight updated and six original Cochrane reviews were identified for the first cohort of reviews, and four co-published Cochrane reviews were included in the second cohort. The original versions of Cochrane Methodology Reviews continued to be cited after the update was published but the median for the total number of citations was non-significantly higher for the updated Cochrane reviews than for their original version after the publication of the updated review [161 (Interquartile range (IQR) 85, 198) versus 113 (IQR 15, 433), p>0.05]. The difference was mainly due to citations in the fourth and fifth year after publication of the update [21 (IQR 10, 49) versus 13 (IQR 4, 53), 30 (IQR 10, 34) versus 10 (IQR 5, 42), all p>0.05], with the median number of citations being lower for the update than for the original version in the first year [2 (IQR 0, 4) versus 20 (IQR 8, 51)] and in the second year [14 (IQR 7, 21) versus 24 (IQR 6, 50)] and similar in the third year [18 (IQR 9, 35) versus 19 (IQR 6, 63)]. All the differences were not statistically significant.
The median number of total citations [362 (IQR 179, 840) versus 145 (IQR 75, 445)] and the median number of citations to the Cochrane review in the first five years after co-publication combined and in each of those years was higher in the co-published group (including both the original review and the co-published version) than in the non-co-published group [first year: 6 (IQR 1, 23) versus 1 (IQR 0, 3); second year: 29 (IQR 16, 63) versus 13 (IQR 5, 17); third year: 36 (IQR 22, 90) versus 19 (IQR 9, 34); fourth year: 37 (IQR 23, 82) versus 22 (IQR 10, 34); fifth year: 59 (IQR 30, 99) versus 15 (IQR 8, 32); sum of years 1-5: 177 (IQR 99, 338) versus 75 (IQR 37, 126)].
The analysis of the adjusted IF showed that, if they had not co-published the Cochrane review, one of the three journals that co-published a Cochrane Methodology Review would have had a lower IF in the first year after the co-publication and two would have had a lower IF in the second year after the co-publication.
Conclusions Earlier versions of Cochrane Methodology Reviews continue to be cited after an update is published, which raises doubts about whether those citing these reviews are using the most recent evidence or are aware of the update. Co-publication is associated with a higher overall citation frequency for Cochrane Methodology Reviews and may increase the IF of the journal in which it is co-published, which facilitates broader application of Cochrane evidence on the methodology of research and promotes its dissemination.