• Amir Reza Amiri, Kavitha Kanesalingam, Suzie Cro, and Adrian T H Casey.
• Spinal Injury Unit, Royal National Orthopaedic Hospital, Brockley Hill, Stanmore, Middlesex, HA7 4LP, United Kingdom. Electronic address: amir.r.amiri@googlemail.com.
• Spine J. 2013 Sep 1;13(9):1148-53.

BackgroundOver the past two decades, there has been a growing recognition and emphasis on the practice of evidence-based medicine (EBM). The level of evidence (LOE) is used to classify clinical studies based on their quality and design. To compare the quality of scientific journals, the impact factor (IF) is the most widely used ranking measure. However, the calculation of IF is not directly dependent on the quality or LOE of clinical articles published in a journal.PurposeThe primary aim of this study was to evaluate the current LOE for clinical research in leading spinal journals and assess the relationship between LOE and IF. We hypothesized that most clinical research would provide level IV evidence, and that a positive correlation would exist between the proportion of high LOE articles and the journal IF.Study DesignA systematic review of all the articles in five general spinal journals was undertaken during 2010.SampleAll online articles in The Spine Journal, Spine, European Spine Journal, Journal of Neurosurgery: Spine, and Journal of Spinal Disorders and Techniques during 2010, as well as supplements were included.Outcome MeasureThe LOE for each clinical study was assessed using guidelines produced by the Oxford Centre for Evidence-Based Medicine.MethodsTwo reviewers independently assessed all articles.ResultsOverall 703 articles were suitable for LOE grading. Of these, 4.7% provided level I evidence, 23.2% level II, 12.5% level III, and 59.6% level IV. There was a significant association between LOE and type of study (p<.001); articles on therapeutic studies had the largest proportion (71.8%) of level IV evidence. There was a strong positive correlation between the proportion of level I and II evidence and the journal impact factor (ρ=0.9; 95% confidence interval 0.1 to 0.99; p=.037).ConclusionSpinal surgery journals with a higher IF contain a larger proportion of studies with high LOE, however most clinical articles provide level IV evidence of which the highest proportion are therapeutic studies. Clinicians, researchers, and journal editors should work hand in hand to enhance evidence-based practice in spinal care.Copyright © 2013 Elsevier Inc. All rights reserved.

13 keywords...

hide…