An international survey indicated that unpublished systematic reviews exist.

Conference item

OBJECTIVE: To determine the frequency of unpublished systematic reviews (SRs) and explore factors contributing to their occurrence. STUDY DESIGN AND SETTING: First or corresponding authors from a sample of SRs published in 2005 were asked to participate in a 26-item survey administered through the Internet, facsimile, and postal mail. Outcomes included median and range of published and unpublished SRs, and barriers, facilitators, and reasons for not publishing SRs. Descriptive analyses were performed. RESULTS: 55.7% (348 of 625) of those invited participated, half of which were from Europe and 22.7% were from the United States. Participants reported 1,405 published (median: 2.0, range: 1-150) and 199 unpublished (median: 2.0, range: 1-33) SRs. Lack of time and lack of funding and organizational support were barriers, whereas time availability and self-motivation were facilitators to publishing reviews. For most recent unpublished SRs (n=52), the reasons for not publishing included lack of time (12 of 52, 23.0%), the manuscript being rejected (10 of 52, 19.0%), and operational issues (six of 52, 11.5%). CONCLUSION: Unpublished SRs do exist. Lack of time, funding, and organizational support were consistent reasons for not publishing SRs. Statistical significance of SR results was not reported as being a major barrier or reason for not publishing. Further research on unpublished SRs is warranted.

Authors: Tricco, A; Pham, B; Brehaut, J; Tetroe, J; Cappelli, M

• Publication status: Published
• Host title: Journal of clinical epidemiology
• Volume: 62
• Issue: 6
• Pages: 617-623.e5
• Publication date: 2009-06-01
• Event location: United States
• DOI: 10.1016/j.jclinepi.2008.09.014
• EISSN: 1878-5921
• ISSN: 0895-4356
• Keywords: Peer Review, Research; Humans; Internationality; Publication Bias; Publishing; Evidence-Based Medicine; Review Literature as Topic; Data Collection; Bias (Epidemiology)