Publication of reviews synthesizing child health evidence (PORSCHE): A survey of authors to identify factors associated with publication in Cochrane and non-Cochrane sources

Abstract

Background: Cochrane Child Health maintains a register of child-relevant Cochrane systematic reviews (SRs) to provide a comprehensive source of high-quality evidence. However, a large number of SRs are published outside of The Cochrane Collaboration (Cochrane), impacting the comprehensiveness of the Cochrane Database of Systematic Reviews (CDSR). We surveyed authors who published child-relevant SRs with Cochrane and elsewhere in the medical literature to (1) understand their experiences in preparing and publishing SRs and (2) identify factors influencing choice of publication venue. Methods: We identified SRs published in the CDSR for the most recent complete year prior to our study (2013; n = 145). We searched the medical literature and randomly selected the same number of SRs published the same year. We developed an internet-based survey and contacted the corresponding author of each review via email. Data were analyzed descriptively. Qualitative analysis elicited common themes from open-ended questions. Results: Seventy-six (26%) responded: 41% Cochrane, 42% non-Cochrane, and 17% published in both venues. Among respondents who published their SR in both venues (n = 13), 46% found it easier to publish in a non-Cochrane journal, 15% easier with Cochrane, and 31% similar. Main reasons for conducting SRs with Cochrane (n = 44) were Cochrane's positive reputation (82%) and good impact factor (66%). Among respondents who published their SR in a non-Cochrane journal (n = 32), most frequent reasons for not conducting their SR with Cochrane were time required to follow Cochrane processes (25%), lack of knowledge about how to conduct an SR with Cochrane (19%), administrative processes (16%), and perception that non-Cochrane journals yielded more interest (16%). Among respondents who published their SR in a non-Cochrane journal (n = 32), 78% did not register their review and 22% did not prepare a protocol. Conclusions: Key reasons for publishing in Cochrane are its positive reputation and impact factor. Reasons for publishing in non-Cochrane sources include lack of familiarity or challenges with the Cochrane processes and desire to publish in a source more directly relevant to the topic of interest. End users looking for evidence in the form of SRs need to be aware that there is a vast number of SRs published across the medical literature. Efforts to optimize the identification of SRs in non-Cochrane sources (e.g., through effective labeling or protocol/review registration) and their content will help end users find the necessary synthesized evidence to support clinical practice.