A descriptive study found low prevalence of presumed predatory publications in a subset of Cochrane reviews.

Objectives: To examine the prevalence of presumed predatory publications in Cochrane reviews, which are considered the gold standard.

Study Design And Setting: We selected two Cochrane networks with broad scope: the Musculoskeletal, Oral, Skin and Sensory Network and the Public Health and Health Systems Network. From reviews produced by all review groups in those networks in 2018 and 2019, we extracted included study citations published after 2000. For each citation, we assessed the journal and publisher using an algorithmic process based on characteristics known to be common among predatory publishers. Knowing that predatory status can be fluid and subjective, we scored citations on a spectrum from "reputable" to "presumed predatory" based on publication characteristics available at the time of assessment.

Results: We assessed 6,750 citations from 300 reviews. Of these citations, 5,734 were published by entities widely accepted as reputable, leaving 1,591 for further assessment. We flagged 55 citations as concerning.

Conclusion: Cochrane reviews across diverse topic areas included studies from flagged publishers, although this number is small. Because of this, there is potential for studies from predatory journals to influence the conclusions of systematic reviews. Researchers should stay aware of this potential threat to the quality of reviews.