• Adam S Tunis, Matthew D F McInnes, Ramez Hanna, and Kaisra Esmail.
• Faculty of Medicine, University of Ottawa, Ottawa, Ontario, Canada; Department of Medical Imaging, Ottawa Hospital, 1053 Carling Ave, Room c159, Ottawa, ON, Canada K1Y 4E9.
• Radiology. 2013 Nov 1;269(2):413-26.

PurposeTo evaluate whether completeness of reporting of systematic reviews and meta-analyses in major radiology journals has changed since publication of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement; a secondary objective is to evaluate whether completeness of reporting (ie, PRISMA) is associated with study quality (ie, Assessing the Methodological Quality of Systematic Reviews [AMSTAR]).Materials And MethodsSystematic reviews and meta-analyses published in major radiology journals between January 2007 and December 2011 were identified by searching MEDLINE with the modified Montori method. Studies were reviewed independently by two investigators and assessed for adherence to the AMSTAR and PRISMA checklists. The average results were analyzed to assess for change in mean score before and after PRISMA publication and to assess results over time; a Pearson correlation coefficient was calculated to assess for any association between PRISMA and AMSTAR results.ResultsIncluded were 130 studies from 11 journals. Average PRISMA and AMSTAR results were 21.8 of 27 and 7.2 of 11, respectively. The average result was higher after publication of PRISMA, and PRISMA-reported items were 22.6 of 27 after publication of PRISMA versus 20.9 of 27 before publication of PRISMA; AMSTAR results were 7.7 of 11 after publication of PRISMA versus 6.7 of 11 before publication of PRISMA. There was a strong positive correlation (r = 0.86) between the PRISMA and AMSTAR results. There was high variability between journals. Radiology had the highest PRISMA reported items (24.7 of 27), and American Journal of Neuroradiology had the lowest (19.6 of 27). Two major areas for improvement include study protocol registration and assessment of risk of bias across studies (ie, publication bias).ConclusionIn major radiology journal studies, there was modest improvement in completeness of reporting of systematic reviews and meta-analyses, assessed by PRISMA, which was strongly associated with higher study quality, assessed by AMSTAR.Supplemental Materialhttp://radiology.rsna.org/lookup/suppl/doi:10.1148/radiol.13130273/-/DC1.RSNA, 2013

11 keywords...

hide…